Understand how a study was done by asking about its methods
The methods section is where a paper earns or loses trust, and it is also the densest part to read: design, participants, materials, procedure, measures and analysis packed into a few paragraphs, often with references to other papers for the details. Ask Search+ "What design did this study use, and how were participants assigned?" and then work through the rest of the methods one question at a time. Each answer cites the methods passage it relied on.
Creating an account needs no payment.
Last updated October 2026
How to read a methods section with Search+
- Upload the paper and any protocol
If the study has a published protocol or a supplementary methods file, add that too. Methods sections in journals are often shortened, and the full procedure lives elsewhere.
- Use the standard order as your question list
Ask about design, then participants, then materials or intervention, then procedure, then measures, then analysis. Each question maps to a part of the methods that is easy to skim past.
- Follow references to earlier methods
When the paper says a procedure was "as described previously", add the cited paper to the workspace and ask the same question across both.
Questions to ask about a paper's methods
Is this a randomized experiment, a cohort, case-control or cross-sectional study, a qualitative study or a mixed design?
How were participants allocated to groups, and who was blinded to the allocation?
What exactly happened to participants, in what order, and over what period?
Which instruments or tests measured the outcomes, and do the authors cite evidence that they are valid?
If this is an interview study, how were participants chosen, how were interviews conducted and how were the transcripts analysed?
Is there enough detail here to repeat the study, or does the paper rely on another source?
What to look for in a methods section
Only some designs can support causal claims. Ask what the design is before reading the results, so you know how strong a claim the study can make.
Recruitment and eligibility shape everything downstream. Look for how participants were found, who was excluded and how many declined.
Abstract ideas such as stress or engagement have to be measured somehow. The methods say how, and that choice can explain why studies disagree.
Some studies register their methods in advance. If the paper mentions a registration, ask whether it reports any changes from the plan.
Ask about one study, or ask across every paper in the workspace to see which designs and measures are used where, with a citation to each paper's methods.
Answers carry inline citations to the methods excerpt, so a description of the procedure can be checked against what the authors wrote.
Methods wording you may see, and what to ask next
| Wording you may see | What it usually signals | A follow-up question |
|---|---|---|
| "randomly assigned" or "randomized" | An experimental design | How was the random sequence generated and concealed? |
| "double-blind" | Neither participants nor assessors knew the allocation | Who exactly was blinded, and was blinding checked? |
| "prospective cohort" | Participants followed forward over time | How long was follow-up, and how many were lost? |
| "semi-structured interviews" | A qualitative method with a topic guide | How were themes derived from the transcripts? |
| "as described previously" | Details are in another paper | Which paper is cited, and is it in this workspace? |
| "pre-registered" | Methods were planned and recorded in advance | Do the authors report any deviations from the registration? |
What is the methodology of a research paper?
The methodology of a research paper is its account of how the study was carried out: the design, the participants or materials, the procedures, how variables were measured and how the data was analysed, in enough detail that others can judge and in principle repeat the work.
Methodology questions
Can Search+ explain the methodology of a paper?
Can it tell me if the methods are sound?
What if the methods refer to another paper?
Can it compare methods across several studies?
Does it work for qualitative research papers?
Can I trust the description without reading the paper?
Know how the study was done before you cite it
Start a workspace, upload the paper, and ask about its design, sample and measures one at a time.
Start a workspace